|Mapping the evolution of 'food deserts' in a Canadian city: supermarket accessibility in London, Ontario, 1961-2005.|
|Jump to Full Text|
|PMID: 18423005 Owner: NLM Status: MEDLINE|
|BACKGROUND: A growing body of research suggests that the suburbanization of food retailers in North America and the United Kingdom in recent decades has contributed to the emergence of urban 'food deserts', or disadvantaged areas of cities with relatively poor access to healthy and affordable food. This paper explores the evolution of food deserts in a mid-sized Canadian city (London, Ontario) by using a geographic information system (GIS) to map the precise locations of supermarkets in 1961 and 2005; multiple techniques of network analysis were used to assess changing levels of supermarket access in relation to neighbourhood location, socioeconomic characteristics, and access to public transit.
RESULTS: The findings indicate that residents of inner-city neighbourhoods of low socioeconomic status have the poorest access to supermarkets. Furthermore, spatial inequalities in access to supermarkets have increased over time, particularly in the inner-city neighbourhoods of Central and East London, where distinct urban food deserts now exist.
CONCLUSION: Contrary to recent findings in larger Canadian cities, we conclude that urban food deserts exist in London, Ontario. Policies aimed at improving public health must also recognize the spatial, as well as socioeconomic, inequities with respect to access to healthy and affordable food. Additional research is necessary to better understand how supermarket access influences dietary behaviours and related health outcomes.
|Kristian Larsen; Jason Gilliland|
Related Documents :
|18423005 - Mapping the evolution of 'food deserts' in a canadian city: supermarket accessibility i...
24128975 - Anorexia of aging: a modifiable risk factor for frailty.
23759265 - Food groups and fatty acids associated with self-reported depression: an analysis from ...
23391045 - Diet and nutrient intake in young adults born preterm at very low birth weight.
23891355 - Establishment, operation and development of a donor human milk bank.
22972785 - Lc/esi-ms/ms analysis outlines the different fumonisin patterns produced by f. verticil...
|Type: Journal Article; Research Support, Non-U.S. Gov't Date: 2008-04-18|
|Title: International journal of health geographics Volume: 7 ISSN: 1476-072X ISO Abbreviation: Int J Health Geogr Publication Date: 2008|
|Created Date: 2008-05-20 Completed Date: 2008-06-05 Revised Date: 2013-06-05|
Medline Journal Info:
|Nlm Unique ID: 101152198 Medline TA: Int J Health Geogr Country: England|
|Languages: eng Pagination: 16 Citation Subset: IM|
|The University of Western Ontario, London, ON, N6A 5C2, Canada. firstname.lastname@example.org|
|APA/MLA Format Download EndNote Download BibTex|
Food Industry / economics, trends*
Food Supply / economics*
Marketing / economics, trends
Journal ID (nlm-ta): Int J Health Geogr
Publisher: BioMed Central
Copyright ? 2008 Larsen and Gilliland; licensee BioMed Central Ltd.
open-access: This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Received Day: 6 Month: 12 Year: 2007
Accepted Day: 18 Month: 4 Year: 2008
collection publication date: Year: 2008
Electronic publication date: Day: 18 Month: 4 Year: 2008
Volume: 7First Page: 16 Last Page: 16
Publisher Id: 1476-072X-7-16
PubMed Id: 18423005
|Mapping the evolution of 'food deserts' in a Canadian city: Supermarket accessibility in London, Ontario, 1961?2005|
|Kristian Larsen1||Email: email@example.com|
|Jason Gilliland1||Email: firstname.lastname@example.org|
1The University of Western Ontario, London, ON, N6A 5C2, Canada
A growing body of research suggests that the suburbanization of food retailers in North America and the United Kingdom in recent decades has contributed to the emergence of urban 'food deserts', that is, socially-distressed neighbourhoods with relatively low average household incomes and poor access to healthy food . While more and more large-format supermarkets are erected on suburban lands, smaller grocers in older central-city neighbourhoods seem to be rapidly disappearing, leaving potential food deserts in their wake. This paper explores the historical and geographical evolution of supermarket access in a mid-sized Canadian city: London, Ontario, 1961?2005.
Why examine access to supermarkets? A healthy diet can reduce the risk of many chronic diseases . The majority of these health problems can be linked to a diet with low fruit and vegetable consumption [3,4] and eating large amounts of sugary or high fat foods . National surveys indicate that most Canadians shop for food at a local supermarket, where the widest variety of products can be found at the most competitive prices . While supermarkets also carry unhealthy foods (e.g., chips, soft drinks, and processed foods), these items are more readily available at neighbourhood convenience stores, which are less likely to offer items supportive of a healthy diet. What happens to residents when the only supermarket in a neighbourhood closes? For certain disadvantaged populations who do not have access to a private vehicle, residing in a food desert may have detrimental effects on overall health and quality of life [7,8].
Environmental inequity occurs when a locality has disproportionately many undesirable characteristics and few desirable characteristics . While environmental equity research is almost always focused on the 'undesirables', that is, identifying demographic disparities in exposure to environmental hazards such as air pollution, toxic waste and other LULUs (locally unwanted land uses), our concern in this paper is the equitable distribution of a particularly 'desirable' sub-category of land use which offers easy access to an array of food and household items at competitive prices: the retail supermarket.
Our primary objective in this paper is to use network-based GIS accessibility measures to determine the extent to which food deserts exist in London. In order to meet this objective, the paper will provide answers to three fundamental research questions.
1. Do systematic spatial inequalities in access to supermarkets exist?
2. Have spatial inequalities in access to supermarkets increased or decreased over time?
3. Do systematic socioeconomic inequalities in access to supermarkets exist?
The scholarly contribution of this paper is twofold: first, for spatial equity research, it offers network-based GIS methods which consider two low-cost travel modes (walking and public transit) for determining accessibility; and second, it gives critical contemporary and historical insights into an urban health issue of increasing present-day interest: the accessibility of healthy and affordable food for disadvantaged populations.
Grocery retailing practices in North American cities have undergone many changes over the past century which have largely been driven by prevailing patterns of urban development . The introduction of the automobile has perhaps had the largest impact on shaping cities over the last century, allowing people to move about much further and more freely, fuelling the rapid postwar suburbanization of the population and the deliverers of the goods and services which they consume . Investment in highway infrastructure, particularly in the period immediately following WWII, cemented the dominance of the automobile as the preferred mode of transit. This form of privatized mobility, however, was not democratically distributed; only those with sufficient wealth could afford the luxury of auto-mobility .
The automobile provided the opportunity for many residents to move to the suburbs, but it was a cultural ideal that stimulated the move to the suburbs from the inner-city. North American society views cities as unhealthy places to live , with poor sanitary conditions, high pollution, unsafe and cramped living spaces. The suburbs were seen as the binary opposite providing healthier opportunities for its inhabitants . At first, retailers and offices remained in the city while residents moved to the suburbs; however, by the 1970's many businesses and retailers (including grocery stores) were moving closer to their predominately suburban customer base .
Food retailing, like most other forms of retailing has seen an increase in both store size and the implementation of chains. In the mid 19th century food was typically obtained through small independently owned markets which were an integral part of residential neighbourhoods. By the turn of the 20th century food retailers had begun to organize into chains, most notably the 'The Great Atlantic and Pacific Tea Company' (A&P). Chaining allowed retailers to lower operating costs, increase profits  and use sophisticated location analysis to provide the most rational spatial arrangement of stores for maximal profits . These resources were not available to independent food stores as they could not afford the associated costs. Chaining was followed by the formation of supermarkets in the 1930's, which had a larger selling area and offered a greater selection of goods . Thus, with stores growing larger in size and investment in location analysis, it was inevitable that the overall number of grocery stores per capita would decrease, and retailers would likely follow their wealthier customers to the suburbs.
Most recent to emerge in the urban food retailing landscape is the giant grocery 'superstore', which can be defined as: a single level store with at least 25 000 square feet of sales area, which sells an array of different foods and household items, and includes a very large parking lot . The establishment of giant superstores on suburban lands meets the physical needs for store development as there is more land available for parking, it is easier for trucks to load and unload, these lands are usually more accessible to highways, and they allow for the development of much larger stores . While superstores are often cheaper to construct in suburban neighbourhoods, they still have created new challenges for planners and engineers around traffic, parking, noise attenuation, public transit and other related issues .
The relocation of supermarkets has been associated with urban development and the physical need for more space, but economic interests were of course the driving force and the fact that major chain stores, from research and experience, typically know where to locate to maximize profit. The location of supermarkets has always been within close proximity to customers with money , while older neighbourhoods have filtered to lower socioeconomic status, the suburbs have typically become more affluent areas; as wealthy residents left the city in favour of suburban living, the supermarkets followed . The effect of the economies of scale parallels the movement of much larger stores as they allow operations to become more efficient . Superstores were able to increase profit as they had lower operating costs, a greater product range, and larger catchment-areas, creating spatial monopolies within their region . Larger stores also meet the social needs of its residents; they have the shelf space to carry many multicultural food items demanded by today's postmodern society and allow for 'one-stop' shopping, as people can find numerous household items along with groceries under one roof, saving valuable time .
The rise of the suburban superstore and abandonment of smaller inner-city supermarkets has also presented challenges for planners and public health policy-makers due to the uneven distribution of healthy, affordable food opportunities [7,18-24]. Several U.S. studies have discovered food deserts in older urban neighbourhoods of low income and high Hispanic or African-American populations [25-27]. On the other hand, recent studies of Canadian cities suggest that urban food deserts are not a problem in Canada. A study of the city of Edmonton, Alberta, found that low-income neighbourhoods near the city centre actually had the best supermarket access . Furthermore, a recent article in the International Journal of Health Geographics by Apparicio and colleagues found that food deserts are "missing" in Montreal and access to healthy food is not a major issue for low income urban residents . Meanwhile, results of studies of food access in the United Kingdom are mixed, suggesting that no clear relationship exists between supermarket access and variables such as location, income, or race in the cities of the U.K. [23,30,31].
The current literature is inconclusive as to whether easier access to healthy/unhealthy foods influences one's overall diet; however, it is widely acknowledged that a healthy diet can reduce the risk of many chronic diseases, including type 2 diabetes, heart disease and certain types of cancer . The majority of these health issues are associated with a diet low in fruit and vegetable consumption [3,4] and high in both fats and sugar . Two American studies have found that African-American residents were more likely to eat a healthy diet when they had access to a local supermarket compared to smaller grocery retailers [27,32].
Research has also determined that when living in a food desert, residents must pay higher prices for groceries at small food shops and convenience stores [33-40]. Studies in London, Ontario and the nearby Waterloo Region have found that residents will have to pay an average of 1.6 times more for identical food items purchased at area convenience stores versus area supermarkets [41,42]. Furthermore, certain disadvantaged populations, including the elderly, disabled, unemployed, and lone-parent households, would be particularly vulnerable to the limited options in a food desert, as a function of low income and restricted mobility [7,43,44].
In North American cities today, the reality is that most new grocery superstores are found, along with other 'big box' outlets, in expansive retail centers which are almost always built in excess of a 500 metre walk of residential land uses, which essentially makes them accessible only to consumers with automobiles . While planners have devoted a great deal of attention in recent years to identifying the "walkability" of different built environments in relation to physical activity, and by extension obesity [45-50], critical dimensions of food environments have not received the same degree of attention [51,52].
Supermarket accessibility is measured in relation to two low-cost modes of intra-urban travel: walking and public transit. Although previous U.S. studies have demonstrated that even people living in poor neighbourhoods often drive to get groceries , we argue that in Canadian cities such as London, Ontario, many people, especially those in lower-income neighbourhoods, must walk or use transit to get groceries. Although the Canadian Census does not include data on automobile ownership, a recent study indicated that only 54% of households in the four most socially-distressed neighbourhoods in London had access to a private automobile .
A spatially-referenced database was created to explore certain historical and geographical dimensions of food deserts for every supermarket in the city in 2005 (n = 28) and 1961 (n = 25). Supermarket addresses were gathered from local business directories, [54,55] and verified using several additional sources: telephone directories (yellow pages); company websites; phone calls to retailers; inspection of air photos, maps, plans; and site visits. The data was geocoded within GIS (ArcGIS 9.2, ESRI) and manually reviewed to ensure extremely high precision: to the front door of each supermarket.
The year 1961 was selected for the beginning of the study, as it was the oldest Census year for which local business directories clearly identified locations of businesses by type . More importantly, four decades was an ideal period of time to observe changes due to the process of suburbanization. The City of London undertook major annexations at the urban/rural fringe during the mid-1960s, which was followed by large-scale suburbanization of businesses and residences. The year 2005 was the most current data available at the time of this study .
GIS-based techniques were used to determine the accessibility of supermarkets, or the relative ease with which Londoners can reach a supermarket on foot or public transit. In order to detect potential spatial variations in accessibility, the city was classified into three areas: urban, suburban, and rural. For this study of urban food deserts, rural areas were excluded. To distinguish which parts of London were urban, suburban, and rural we scrutinized phases of historical urban development, particularly how the official city limits expanded over time. Following the practice of local planners, we considered the area within the limits of the City of London as of 1959 to be urban, the area annexed to the City between 1960 and 1992 as suburban, and the lands added since 1993 as rural. Urban and suburban areas were examined in detail at the census tract (CT) level. Census tracts were used as the geographic unit of analysis, as they are commonly-used proxies for 'neighbourhoods' and they are the smallest area unit available that has all the relevant data necessary for the study. In 2001 (the latest census year for which data are available), there were a total of 76 residential CTs within urban and suburban London.
Typical analysis conducted in public health and environmental justice research uses either a 'container' or a 'buffer' approach [56-61]. The container method  does not precisely examine proximity; rather, it identifies whether selected features (e.g., hazards) are contained within, or coincide with, a chosen area unit (e.g., county, city, census tract). As such, this method has also been described as 'spatial coincidence'  and 'unit-hazard coincidence' . While some previous studies have used a circular 'buffer' to identify all sites located within a pre-determined, straight-line distance (or ring) from origin, the circular buffer technique does not take into consideration how variations in the configuration of the circulation networks along which people actually move can affect levels of access [42,62]. In order to overcome the shortfalls of the classic container or buffer approach, we adopted a network-based approach using the city's circulation system (e.g., streets, footpaths) and public transit routes (bus lines) in calculating the distance one can travel.
Accessibility by foot has typically been measured according to a pre-defined distance from origin (i.e., home). In older food desert studies, a distance of 500 metres was commonly used to assess accessibility by walking [63-65], whereas more recently, two Canadian studies use a distance of 1000 metres [28,29] (or a 10?15 minute walk) to represent accessibility. We originally used both 500 metre and 1000 metre distances in our analyses , but for sake of brevity and comparison with other Canadian studies, we will report the results for 1000 metre distances only (additional tables available upon request). Network analysis was employed to create a 'service area' of 1000 metres around each supermarket using the London 2005 street network file (obtained from the City of London Planning Department) within ArcGIS 9.2 (ESRI). We used the 2005 street file also for the 1961 analysis as a careful comparison against the Underwriter's Survey Bureau insurance atlas (1958?62) revealed that the street pattern around the 1961 supermarkets had not changed between 1961 and 2005. Figure 1 illustrates the location of supermarkets in 1961 and 2005 along with the 1000 metre service areas and city boundaries used in this study.
Block level population counts were used to identify the population with supermarket access within each service area. Blocks are the smallest geographic unit for which Statistics Canada releases population counts, and are very useful units for precise spatial analysis. (Unfortunately, virtually all other useful data is suppressed at this scale due to privacy policies). The use of block level data is very precise as it accounts for micro-level variations in population density, rather than assuming that the population is evenly distributed. The supermarket service area were divided into sections and assigned to the appropriate CT for analysis at the census tract level. If a block centroid (i.e., the geographic center) fell within a supermarket service area, the population of that block was identified as having supermarket access. This method allows us to determine how many people live within a walkable distance to a supermarket.
Public transit is typically the only option disadvantaged populations have to reach destinations beyond walking distance. We therefore also included data on London's public transit network in our analysis to determine accessibility by city bus. Bus access was determined by using a 10-minute bus ride without transfers, combined with a 500-metre walk at the beginning and/or end of the bus trip. Figure 2 displays the location of bus routes and supermarkets within the city of London. As expected most bus routes permeate inner-city neighbourhoods.
Bus schedules and route data of the London Transit Commission (LTC) were used to determine that a 10-minute bus ride covers a trip of 3 kilometres. Network Analyst was used to select and map individual bus routes that are located within 500 metres and extend up to 3 kilometres from each supermarket; a 500-metre network service area was created around each identified bus route to determine areas with public transit access to supermarkets. Public transit is an option often ignored in previous food desert studies and allows us to more accurately represent supermarket access for London residents.
The distance to the closest supermarket was computed using the street network file and the Network Analyst extension within ArcGIS 9.2 to determine proximity or the minimum distance residents must walk to the closest supermarket. This measure calculates the shortest path along the street network from every block centroid (within the urbanized area) to the closest supermarket. The data were then aggregated to the census tract level for further analysis and to allow comparison with socioeconomic characteristics of neighbourhoods, in the same manner as previous Canadian studies (equation 1) [28,29].
(1) mean distance to nearest supermarket (census tract) = ?tp (min|dps|)/?tp
p = block population
s = supermarket
tp = total block population within entire census tract
dps = distance between block centroid and supermarket
Network Analyst was also employed to calculate the number of supermarkets within 1000-metres of each block centroid. This allows researchers to determine diversity or how many supermarkets residents have access to by foot. Once again the data were aggregated to the census tract level (equation 2) in order to compare with other socioeconomic datasets.
(2) mean number of supermarkets within 1000 metres (census tract) = (?tp ?sc)/?tp
p = block population
c = census tract
s = supermarkets
sc = number of supermarkets within 1000 metres of block centroid
tp = total block population within entire census tract
To test hypotheses regarding neighbourhood socioeconomic characteristics and supermarket accessibility we first devised a strategy for identifying the relative socioeconomic status (SES) of each CT. Following previous studies of 'disadvantaged', 'deprived' or 'underclass' neighbourhoods [57,60,66-68], census tracts were characterized by a composite index of socioeconomic distress comprised of four variables drawn from the 2001 Canadian census: low educational attainment (proportion of individuals that have not graduated from high school), lone parenthood (proportion of lone parent families versus the total number of families), unemployment (unemployment rate), and incidence of low income (proportion of households that fall below the low income cut-off according to Statistics Canada). Each individual variable was considered on a one-by-one basis, then together in a composite index. Although each of the four variables is positively correlated with the other three (Table 1), we incorporated them all into one composite index in order to highlight the neighbourhoods with multiple indicators of distress . Following the method used by Gilliland and Ross , we first calculated the z-scores (based upon un-weighted mean and standard deviation of the indicators) of each CT for each of the four indicators, and then summed them to get a total distress index score for each CT. The census tracts were then classified into three categories (low distress, moderate distress, and high distress) for comparison and analysis.
Spatial inequalities in access to supermarkets were explored in urban versus suburban London; the location of all 28 supermarkets with 1000-metre network-based service areas is displayed in Figure 1A. Initial analysis indicates that several regions of the city appear to have limited access to supermarkets, and therefore may be considered 'food deserts' (Figure 1) [69,70]. Several large areas of the suburbs were not within walking distance of a supermarket and there was also a large food desert covering the city centre and areas immediately east, in neighbourhoods known locally as 'Central London' and 'East London'.
Table 2 displays the average proportion of neighbourhood population with supermarket access based on walking and transit access. In both measures the urban areas had better access than the suburban neighbourhoods.
The city of London has gone through dramatic changes over the past four decades and the city's population has nearly doubled (population 1961: 169 569; population 2005: 325 045) [69,70]. To establish whether spatial inequalities in access to supermarkets have changed over time, we examined London's supermarket locations in 1961 and compared the population served by each of the major grocery retailers in 1961 (n = 25) and 2005 (n = 28). Figure 1 illustrates changes in supermarket location and access from 1961 to 2005. Analysis indicates that supermarket access has diminished over time, as the average proportion of census tract population with easy supermarket access in 1961 (45.2%) was more than twice the level for 2005 (18.3%) [69,70].
The number of people in each CT with easy supermarket access in 1961 and 2005 was measured using 1000-metre network-based service areas, and a comparison of the maps in Figure 3 reveals that people in the central core of the city in 1961 were much better served than they are today. In 1961, the majority of CTs in the core of the city had greater than 2500 residents within easy walking distance of a supermarket (Figure 3A); whereas in 2005, only a third of urban CTs had greater that 2500 residents within walking distance, and a third of CTs did not have any residents living within walking distance (Figure 3B). When the proportion of residents with access is considered (Figure 3C; Figure 3D) similar results are discovered. The majority of census tracts in central areas of the city for 1961 had greater than 75% of the population with access to supermarkets, while in 2005 most urban areas displayed that less than 20% of residents had access.
To explore socioeconomic inequalities in relation to supermarket access we first characterized each neighbourhood by a composite index of socioeconomic distress. The results reported in Table 3 indicate that the most distressed areas of the city had the poorest levels of access to supermarkets by walking, while the least distressed had the best access. When public transit is considered, the most distressed areas of the city had significantly improved access to supermarkets, and the middle class areas had the lowest levels of access, a result of the fact that most bus routes permeate low income urban neighbourhoods.
Figure 4A maps the percentage of CT population with supermarket access by foot as graduated circles against a thematic map of CTs shaded according to socioeconomic distress index score. For each CT, larger circles indicate increasing levels of access and darker shading represents higher levels of socioeconomic distress. The results suggest that the most distressed urban CTs have the poorest levels of access to supermarkets. On the other hand, CTs with the lowest levels of socioeconomic distress (suburban areas in the north and west of the city) also have relatively low levels of access to supermarkets; middle class areas appear to be the best served.
If we consider the availability of public transit in our analysis of accessibility, the least distressed neighbourhoods appear to have the poorest access to supermarkets, followed by the most distressed areas. Figure 4B) displays access by public transit in relation to neighbourhood distress level. Figure 4B) again demonstrates that CTs in the highly-distressed East London neighbourhood have the poorest levels of access to supermarkets, while middle-class CTs have the greatest levels of access.
When we measure access by the number of supermarkets within 1000 metres of each block centroid (aggregated to CT), we do not find significant differences between supermarket access and level of distress, as the figures only varied by 0.07 (Table 4). Figure 5A illustrates the number of supermarkets within 1000 metres, and again the East London region has poorest access to supermarkets.
Our measurements of distance to the nearest supermarket suggest that there is no significant difference in supermarket access related to neighbourhood socioeconomic distress level. Residents in areas of the highest socioeconomic distress had the furthest to travel at 1346 metres (Table 4), but this is only about 66 metres further than residents in the best-served areas of moderate distress levels. Figure 5B illustrates the poor access to grocers in East London, where distance to the nearest supermarket is over 1600 metres.
While no significant correlation was found between level of supermarket access and presence of low income (Table 1); the East London neighbourhood is still believed to be a food desert. With all four measures of analysis East London was always in the category representing poorest access to supermarkets.
The suburbanization of grocery stores over the last four decades in London has created a distinct 'food desert' in the central and eastern sections of the urban core. In 1961, the vast majority of grocers in London were located in the inner city and very few were in the suburbs; however, the landscape of grocery retailing is very different today, as supermarkets have left the older, inner-city neighbourhoods such as Central London and East London, and these low-income areas now have the poorest levels of supermarket access by walking. These changes in retailing practices are also driven by transformations in demographics within the city. In 1961, the majority of the population lived in central areas near downtown, while in 2005 these areas near the core have lost population and are now even less dense than a few suburban neighbourhoods. Not surprisingly, the new supermarkets opening in suburban areas have been designed for the car-borne suburban consumer. In fact, the massive parking lots that surround the so-called 'superstores' today have made the present-day grocery shopping trip virtually impossible for pedestrians. The footprint of the typical supermarket in London has increased dramatically from a mere 850 square metres of floor area in 1961  to a whopping 4000 square metres today . These changes in supermarket location and floor area can be attributed to the changing business practices of grocery retailers nationwide.
Since it is somewhat unrealistic to think that everyone should live within walking distance to a supermarket, we also assessed access via public transit. As the majority of the bus lines go through the urban core, we were not surprised to find that most urban census tracts had superior access by transit; however, it is noteworthy that the highly-distressed neighbourhood of East London also had below average level of access by bus. As suggested in previous work , we can assume that people in poorer, more distressed, neighbourhoods do not have as high a level of access to private automobiles as people in less distressed, high income areas, and are therefore more likely to walk or rely on public transit to shop for groceries. When walking is not feasible and transit access is not available, residents may use a taxi for transportation, but taxi fare adds a significant cost to the household budget which can be troublesome for persons of low income or fixed incomes. Furthermore, taxi dependency would also necessitate making fewer trips to the supermarket to avoid multiple fares; and, in turn, less frequent trips would be detrimental to the household's ability to take full advantage of time-sensitive sales, to keep fresh produce in the home, and to minimize the total financial outlay for any one shopping trip.
Based on our results with respect to the measures of distance to nearest supermarket and number of supermarkets within 1000 metres, it can be concluded that supermarket accessibility is poor throughout the city of London. Furthermore, according to these measures, supermarket access varied little in relation to level of socioeconomic distress, as each category displayed relatively dismal access. While supermarket access is poor throughout the city of London, the overall findings indicate that distinct food deserts do exist, particularly in the East London neighbourhoods.
Our findings differ in comparison to another Canadian study in this journal by Apparicio and colleagues , who found that most areas of the city of Montreal had good access to supermarkets, except for suburban lands with low distress, leading them to claim that food deserts do not exist as a food desert can only be present when there is both poor access and high distress. Although they are both Canadian cities, Montreal and London are very different in terms of population and urban form. Montreal is much larger than London in terms of total population, at 1.8 million versus 350,000 inhabitants, and it also has a much higher population density: 847 versus 185 persons per square kilometre . While food deserts are not always a function of population density, retail businesses such as grocers require a certain customer base (i.e., population density threshold) to be profitable; it is not surprising that London's lower population density is associated with a more dispersed distribution of supermarkets than in higher-density Montreal. Furthermore, London's low-density form makes the city much less 'walkable' than Montreal. This lack of accessibility makes automobile ownership a practical necessity for many routine daily functions such as grocery shopping.
Londoners living in inner-city neighbourhoods of low socioeconomic status have very poor levels of access to supermarkets by foot; indeed, there are no supermarkets in neighbourhoods that locals would associate with 'downtown living'. Ironically, the high income neighbourhoods with the lowest level of socioeconomic distress also had relatively poor supermarket access. This finding is perhaps not surprising, as every census tract in the lowest social distress category was located in a newer suburb of London, where automobiles are a necessity for everyday life. Additionally, the median household income for these suburban areas was typically much higher than other regions of the city: $91,172, versus the city median of $48,026 and $36,583 for East London . In London, as in many North American cities, households of higher income generally reside in larger houses on larger lots, in neighbourhoods of relatively low density, that tend to be zoned exclusively residential; thus, the physical design of the suburban environment dictates that these households will not have good access to supermarkets within walking distance.
While this study makes methodological advances by incorporating GIS-based techniques for evaluating accessibility by foot and public transit in an urban setting, an obvious limitation is the highly empirical nature of the study. The most useful avenue for future research would be to undertake interviews with people who live in food deserts in order to gain a better understanding of the personal, psychological, economic and geographical dimensions of grocery shopping patterns and behaviours. As census data on automobile ownership in London does not currently exist, surveys and interviews are needed to determine levels of vehicle ownership in different neighbourhoods, and how (lack of) access to a personal vehicle influences grocery shopping behaviours.
Conclusions as to whether easy access to supermarkets improves one's diet and overall quality of life have also been mixed [32,73-76]. Residents of a food desert may not have access to a supermarket, but other types of small food retailers such as a local butcher, fruit and vegetable market, baker, or ethnic and speciality food stores may exist. Local accessibility to these kinds of smaller food stores may improve residents' access to healthy foods. Although this type of geographical access may exist, food type, quality, and price should also be examined to help us understand diet and consumer costs of life in a food desert in London, Ontario (and elsewhere).
It can be argued that urban food deserts are merely another product of the social and economic forces underlying the historical evolution of urban form. Basically, new supermarkets are being erected at the suburban fringe and older ones are being shutdown in the inner-city because it makes economic sense for the owners of the supermarket chains, who are following the suburbanization of their customer base. Furthermore, in suburbia, the grocery chains are able to access larger tracts of developable land, often at a much cheaper price than in urban areas, in order to build giant superstores with ample parking. Studies have also found that operating costs in urban locations are typically higher than in suburbs . The bottom line is that major grocery retailers know where to locate a supermarket to maximize profit and are likely to continue the trend of building superstores in the suburbs as long as it is the most profitable solution.
The reintroduction of small supermarkets in Central or East London could have many benefits to not only the citizens currently living in these areas, but it could also help attract more people to the downtown core. London City Councillors have publicly acknowledged the need for a supermarket in the core districts in order to attract and maintain residents and businesses . In order to attract small supermarkets back to the urban core, the municipal government could introduce financial incentives, such as tax breaks or building restoration initiatives. The City could also change zoning or parking regulations in certain areas to encourage grocery chains to consider locations in need. Such incentive programs and regulatory modifications are in line with the mandates of local community organizations Mainstreet London and the Old East Village Business Improvement Area that are working to bolster economic development in Central and East London respectively [78,79]. In other major cities across Canada (e.g., Toronto, Montreal, Edmonton), supermarkets can still be found in inner-city locations, and new ones are continually being built; but again, these cities are already larger, with higher population densities in their central cores. The bottom-line is that people need supermarkets (the political/popular view), and supermarkets need people (the business view); any municipal strategy for inner-city development or revitalization must recognize the positive correlation between supermarkets and population density. Regardless of the chosen strategy, the City of London should actively encourage supermarket development in areas currently identified as food deserts.
This study explored issues of environmental equity with respect to the historical evolution of food deserts in London, Ontario. The findings confirmed that some areas of the city currently have significantly better access than others, and the East London neighbourhood is indeed a food desert. Historical analysis, however, indicated that inner-city neighbourhoods such as East London have not always been food deserts. The research suggests how food deserts are the result of structural and geographical changes in the business of grocery retailing over the past few decades, as the majority of supermarkets that existed in the city in the 1960s have since vacated their original urban locations, and the majority of new stores have been built in the suburbs. This study is the first known attempt to present a historical analysis of the spatial patterning of supermarket access in a mid-sized North American city, thereby contributing important new information to our understanding of the evolution of urban food deserts.
A major contribution of this research resulted from the utilization and evaluation of different measures to determine accessibility; the methods outlined here should serve as guidelines for future studies of environmental equity. To further expand the knowledge base on food deserts, access to public transit was assessed to more accurately measure how low-income residents can realistically access supermarkets. This consideration added a very important dimension to this study, as previous studies have failed to examine access via public transit. This failing of previous studies is even more significant considering that most researchers have focussed on larger cities, which tend to have more advanced public transit networks (including subway systems). Our results found that populations in the majority of the urban census tracts had very good access to supermarkets via public transit, but the population in East London still had poorer access by transit, compounding the impacts of the food desert.
While the current retailing practice of building supermarkets only on 'greenfields' at the suburban/rural fringe does not seem to show signs of slowing, this study demonstrates some of the negative consequences of this trend. This practice causes many smaller urban supermarkets to close, and food deserts to form throughout the city. Governments at all levels must evaluate and react to the various consequences associated with the closure of smaller urban supermarkets. Ultimately, the closure of the supermarkets will result in an increased unemployment rate in distressed cores, fewer visitors to surrounding retailers, and potential impacts to the health and well-being of already vulnerable populations.
The authors declare that they have no competing interests.
KL completed the GIS-based analyses and wrote a draft of the manuscript. JG designed the study, edited and wrote sections of the paper. Both KL and JG performed the analyses. All authors read and approved the final version of the manuscript.
This research is supported by the Canadian Institutes for Health Research (PI: Jason Gilliland). Certain data has been provided by the City of London Planning Department. The authors would also like to thank Mathew Novak (UWO-Geography), Kevin Van Lierop (UWO), Dr Maged Boulos (Editor), and the four peer reviewers for their insightful comments which have helped to improve this paper.
|Reisig V,Hobbiss A. Food deserts and how to tackle item: A study of one city's approachHealth Education Journal 2000;59:137–149. [doi: 10.1177/001789690005900203]|
|Health CanadaNutrition and healthy eating. Assessed November 20, 2007.|
|Darmon N,Ferguson E,Briend A. A cost constraint alone has adverse effects on food selection and nutrition density: An analysis of human diets by linear programmingJ Nutr 2002;132:3764–3771. [pmid: 12468621]|
|Kyle R,Blair A. Planning for health: Generation, regeneration and food in SandwellInt J Retail Distrib Manag 2007;35:457–473. [doi: 10.1108/09590550710750331]|
|Must A,Spadano J,Coakley EH,Field AE,Colditz G,Dietz WH. The disease burden associated with overweight and obesityJ Am Med Assoc 1991;282:1523–1529. [doi: 10.1001/jama.282.16.1523]|
|Health CanadaCanada's food guide to healthy eating. Accessed November 3, 2005.|
|Kirkup M,Kervenoael RD,Hallsworth A,Clarke I,Jackson P,Aguila RPd. Inequalities in retail choice: Exploring consumer experiences in suburban neighbourhoodsInt J Retail Distrib Manag 2004;32:511–522. [doi: 10.1108/09590550410564746]|
|Lake A,Townshend T. Obesogenic environments: Exploring the built and food environments J R Soc Health 2006;126:262–267. [pmid: 17152319] [doi: 10.1177/1466424006070487]|
|Greenberg M,Cidon M. Broadening the definition of environmental equity: A framework for states and local governmentsPop Res Policy Rev 1997;16:394–413.|
|Pothukuchi K. Attracting supermarkets to inner-city neighbourhoods: Economic development outside the boxEcon Dev Q 2005;19:232–244. [doi: 10.1177/0891242404273517]|
|Duany A,Plater-Zyberk E,Speck J. Suburban Nation: The Rise of Sprawl and the Decline of the American Dream. 2001New York: North Point Press;|
|Jackson K. Crabgrass Frontier: The Suburbanization of the United States. 1985New York: Oxford University Press;|
|Boone CG,Modarres A. City and Environment. 2006Philadelphia, PA: Temple University Press;|
|Wrigley N,Lowe M. Reading Retail: A Geographical Perspective on Retailing and Consumption Spaces. 2002New York: Oxford University Press;|
|Jones KG,Simmons JW. Location: Analysing the Retail Environment. 1993Scarborough, ON: Nelson Canada;|
|Boarnet MG,Randall C,Chatman DG,Manville M. Emerging planning challenges in retailJ Am Plann Assoc 2005;71:433–449. [doi: 10.1080/01944360508976714]|
|Donohue RM. Abandonment and Revitalization of Central City Retailing: The Case of Grocery Stores. Unpublished doctoral dissertation, University of Michigan, Ann Arbor;|
|Clifton KJ. Mobility strategies and food shopping for low-income familiesJ Plann Educ Res 2004;23:402–413. [doi: 10.1177/0739456X04264919]|
|White M. Food access and obesityObes Rev 2007;8:99–107. [pmid: 17316311] [doi: 10.1111/j.1467-789X.2007.00327.x]|
|Guy CM. The Retail Development Process. 1994London, UK: Routledge;|
|Guy CM. Corporate strategies in food retailing and their local impacts: A case study of CardiffEnviron Plan A 1996;28:1575–1602. [doi: 10.1068/a281575]|
|Wrigley N. Understanding store development programmes in post-property-crisis UK food retailingEnviron Plan A 1998;30:15–35. [doi: 10.1068/a300015]|
|Cummins S,Macintyre S. The location of food stores in urban areas: A case study in GlasgowBr Food J 1999;101:545–553. [doi: 10.1108/00070709910279027]|
|Clarke G,Eyre H,Cliff G. Deriving indicators of access to food retail provision in British cities: Studies of Cardiff, Leeds and BradfordUrban Stud 2002;39:2041–2060. [doi: 10.1080/0042098022000011353]|
|Curtis KA,McClellan S. Falling through the safety net: Poverty, food assistance and shopping constraints in an American cityUrban Anthropol Stud Cultur Syst World Econ Dev 1995;24:93–135.|
|Weinberg Y. No place to shop: Food access lacking in the inner cityRace Poverty Environ 2000:22–24.|
|Morland K,Wing S,Roux AD. The contextual effects of the local food environment on residents' diets: The atheroscelerosis risk in communities studyAm J Public Health 2002;92:1761–1767. [pmid: 12406805]|
|Smoyer-Tomic KE,Spence JC,Amrhein C. Food deserts in the prairies? Supermarket accessibility and neighbourhood need in Edmonton, CanadaProf Geographer 2006;58:307–326. [doi: 10.1111/j.1467-9272.2006.00570.x]|
|Apparicio P,Cloutier MS,Shearmur R. The case of Montreal's missing food deserts: Evaluation of accessibility to food supermarketsInt J Health Geogr 2007;6 [pmid: 17295912]|
|Cummins S,Macintyre S. "Food deserts" ? evidence and assumption in health policy makingBMJ 2002;325:436–438. [pmid: 12193363] [doi: 10.1136/bmj.325.7361.436]|
|Cummins S,Petticrew M,Sparks L,Findlay A. Large scale food retail interventions and dietBMJ 2005;330:683–684. [pmid: 15790619] [doi: 10.1136/bmj.330.7493.683]|
|Zenk S,Schultz A,Israel B,James S,Bao S,Wilson M. Neighbourhood racial composition, neighborhood poverty, and the spatial accessibility of supermarkets in metropolitan DetroitAm J Public Health 2005;95:660–667. [pmid: 15798127] [doi: 10.2105/AJPH.2004.042150]|
|Horton S,Campbell C. Do the poor pay more for food?Food Market Commen 1990;11:33–38.|
|Sooman A,Macintyre S,Anderson A. Scotland's health: A more difficult challenge for some? The price and availability of healthy foods in socially contrasting localities in the west of ScotlandHealth Bull (Edinb) 1993;51:276–284. [pmid: 8225953]|
|Travers K. The social organization of nutritional inequitiesSoc Sci Med 1996;43:543–553. [pmid: 8844955] [doi: 10.1016/0277-9536(95)00436-X]|
|Alwitt L,Donley T. Retail stores in poor urban neighbourhoodsJ Consum Aff 1997;31:139–164.|
|Barratt J. The cost and availability of healthy food choices in southern DerbyshireJ Hum Nutr Dietetics 1997;10:63–69. [doi: 10.1046/j.1365-277X.1997.00487.x]|
|Caraher M,Dixon P,Lang TR. Barriers to accessing healthy foods: Differentials by gender, social class, income and mode of transportHealth Educ J 1998;57:191–201. [doi: 10.1177/001789699805700302]|
|Kayani N. Food Deserts: A Practical Guide. 1998London, UK: Chadwick House Publishing;|
|Chung C,Myers SL. Do the poor pay more for food? An analysis of grocery store availability and food price disparitiesJ Consum Aff 1999;33:276–296.|
|Region of Waterloo Public HealthA glance at access to food. Assessed December 1, 2005.|
|Larsen K,Gilliland J. Environmental equity and the grocery cart: exploring urban 'food deserts'Presented to the Annual Meeting of the Association of American Geographers Chicago, IL. 2006|
|Whelan A,Wrigley N,Warm D,Cannings E. Life in a 'Food desert'Urban Stud 2002;39:2083–2100. [doi: 10.1080/0042098022000011371]|
|Rex D,Blair A. Unjust des(s)erts: Food retailing and neighbourhood health in SandwellInt J Retail Distrib Manag 2003;31:459–465. [doi: 10.1108/09590550310491423]|
|Hess PM,Vernez Moudon A,Snyder MC,Stanilov K. Site design and pedestrian travelTransportation Res Rec 1999:9–19. [doi: 10.3141/1674-02]|
|Boarnet M,Crane R. Travel by Design: The Influence of Urban Form on Travel. 2001Oxford, UK: Oxford University Press;|
|Handy SL,Boarnet MG,Ewing R,Killingsworth RE. How the built environment affects physical activityAm J Prev Med 2002;23:64–73. [pmid: 12133739] [doi: 10.1016/S0749-3797(02)00475-0]|
|Frank LD,Andresen M,Schmid T. Obesity relationships with community design, physical activity, and time spent in carsAm J Prev Med 2004;27:87–96. [pmid: 15261894] [doi: 10.1016/j.amepre.2004.04.011]|
|Frank LD,Sallis JF,Conway TL,Chapman JE,Saelens BE,Bachman W. Many pathways from land use to health: Associations between neighborhood walkability and active transportation, body mass index, and air qualityJ Am Plann Assoc 2006;72:475–487.|
|Frank LD,Engelke PO,Schmid TL. Health and Community Design: The Impact of the Built Environment on Physical Activity. 2003Washington, DC: Island Press;|
|Boarnet MG. Planning's role in building healthy citiesJ Am Plann Assoc 2006;72:5–9. [doi: 10.1080/01944360608976719]|
|Pothukuchi K,Kaufman J. The food system: A stranger to the planning fieldJ Am Plann Assoc 2000;66:113–124. [doi: 10.1080/01944360008976093]|
|Luciani M. The geography of emergency food assistance in London, Ontario. 2005The University of Western Ontario: Master's Thesis;|
|Vernon's City DirectoryCity of London. 1962Hamilton: Vernon Directories Ltd;|
|Vernon's City DirectoryCity of London. 2005Hamilton: Vernon Directories Ltd;|
|McMaster RB,Leitner H,Sheppard E. GIS-based environmental equity and risk assessment: Methodological problems and prospectsCartography & Geographic Information Systems 1997;24:172–189. [doi: 10.1559/152304097782476933]|
|Gilliland J,Ross N. Opportunities for video lottery terminal gambling in Montr?al: An environmental analysisCan J Public Health 2005;96:55–59. [pmid: 15682698]|
|Mohai P,Saha R. Reassessing racial and socioeconomic disparities in environmental justice researchDemography 2006;43:383–399. [pmid: 16889134] [doi: 10.1353/dem.2006.0017]|
|Wilson DH,Gilliland J,Ross NA,Derevensky J,Gupta R. Video lottery terminal access and gambling among high school students in MontrealCan J Public Health 2006;97:202–206. [pmid: 16827406]|
|Gilliland J,Holmes M,Tucker P,Irwin J. Environmental equity is child's play: Mapping public provision of recreation opportunities in urban neighbourhoodsVulnerable Child Youth Stud 2006;1:256–268. [doi: 10.1080/17450120600914522]|
|Anderton DL,Anderson AB,Oakes JM,Fraser MR. Environmental Equity: The demographics of dumpingDemography 1994;32:229–248. [pmid: 7926187] [doi: 10.2307/2061884]|
|Oliver LN,Shuurman N,Hall AW. Comparing circular and network buffers to examine the influence of land use on walking for leisure and errands Int J Health Geogr 2007;6:41. [pmid: 17883870] [doi: 10.1186/1476-072X-6-41]|
|Donkin A,Dowler E,Stevenson S,Turner S. Mapping access to food at a local levelBr Food J 1999;101:554–564. [doi: 10.1108/00070709910279054]|
|Furey S,Strugnell C,McIlveen H. An investigation of the potential existence of "food deserts" in rural and urban areas of Northern IrelandAgric Hum Values 2001;18:447–457. [doi: 10.1023/A:1015218502547]|
|Wrigley N. 'Food deserts' in British cities: Policy context and research prioritiesUrban Stud 2002;39:2029–2040. [doi: 10.1080/0042098022000011344]|
|Ricketts E,Sawhill I. Defining and measuring the underclassJ Policy Anal Manage 1988;7:316–325. [doi: 10.2307/3323831]|
|Carstairs V,Morris R. Deprivation and Health in Scotland. 1991Aberdeen, UK: Aberdeen University Press;|
|Ley D,Smith H. Relations between deprivation and immigration in large Canadian citiesUrban Stud 2000;37:37–62. [doi: 10.1080/0042098002285]|
|Statistics CanadaCensus of Canada. 1961Ottawa: Dominion Bureau of Statistics;|
|Statistics CanadaCensus of Canada. 2001Ottawa: Statistics Canada;|
|Underwriter's Survey BureauInsurance Atlas for the City of London. 1962Toronto and Montreal: Underwriter's survey bureau;|
|City of LondonMapping data distribution. 2005London, Ontario: Corporation of the City of London, Geomatics Division;|
|Wrigley N,Warm D,Margetts B. Deprivation, diet, and food-retail access: Findings from the Leeds 'food deserts' studyEnviron Plan A 2003;35:151–188. [doi: 10.1068/a35150]|
|Pearson T,Russell J,Campbell MJ,Barker ME. Do 'food deserts' influence fruit and vegetable consumption? A cross-sectional studyAppetite 2005;45:195–197. [pmid: 15927303] [doi: 10.1016/j.appet.2005.04.003]|
|Turrell G,Blakely T,Patterson C,Oldenburg B. A multilevel analysis of socioeconomic (small area) differences in household food purchasing behaviourJ Epidemiol Community Health 2004;58:208–215. [pmid: 14966233] [doi: 10.1136/jech.2003.011031]|
|White M,Bunting J,Raybould S,Adamson A,Williams L,Mathers J. Do food deserts exist? A multi-level, geographical analysis of the relationship between retail food access, socio-economic position and dietary intake. 2004UK, Food Standards Agency;|
|Miner J,Dubinski K,Daniszewski H. Core renewal ideas abound. The London Free Press; March 19, 2007.|
|Old East Village Business Improvement Area Assessed November 20, 2007.|
|Mainstreet London Assessed November 20, 2007.|
Correlation between Level of Neighbourhood Distress and Supermarket Access
|(a)||Lone-parent families (%)||1|
|(b)||Incidence of low income (%)||0.802**||1|
|(c)||Low educational attainment (%)||0.626**||0.587**||1|
|(d)||Unemployment rate (%)||0.464**||0.522**||0.541**||1|
|(f)||Nearest supermarket (m)||0.071||0.153||0.051||-0.003||0.045||1|
|(g)||Number of supermarkets within 1000 m||0.061||0.009||0.020||-0.030||-0.063||-0.685**||1|
|(h)||% Population with access 1000 m network||-0.079||-0.201||-0.014||-0.054||-0.101||-0.828**||0.746**||1|
|(i)||% Population with bus access||0.238*||0.034||-0.003||0.090||0.104||-0.450**||0.278*||0.420**||1|
** Correlation is significant at the 0.01 level (2-tailed).
* Correlation is significant at the 0.05 level (2-tailed).
Proportion of census tract population with supermarket access
|Districts||% population with access 1000 m||% population with bus access*|
* Bus access uses a 10 minute bus ride and 500 metre walk
Proportion of census tract population with supermarket access according to socioeconomic distress, 2005
|Distress||% population with access 1000 m||% population with bus access*|
* Bus access uses a 10 minute bus ride and 500 metre walk
Descriptive statistics for supermarket access, 2005
|Number of supermarkets within 1000 m||Nearest supermarket (metres)||% population with access 1000 m||% population with bus access|
Previous Document: Stimulatory effect of Eucalyptus essential oil on innate cell-mediated immune response.
Next Document: Empowering the people: Development of an HIV peer education model for low literacy rural communities...